Coding agents, in the vendors’ own words
Every cell quotes the vendor. Checked 18 Nov 2026: 126 of 126 quotes still on their pages. How
Coding agents, feature by feature
| MCP servers | ||||||||
|---|---|---|---|---|---|---|---|---|
| Models from several companies | ||||||||
| Edits across files | ||||||||
| Works in the cloud | ||||||||
| Reviews pull requests | ||||||||
| Runs in the terminal | ||||||||
| VS Code and JetBrains | ||||||||
| SSO and SCIM | ||||||||
| Audit logs | ||||||||
| Usage reports | ||||||||
| Data residency | ||||||||
| Not used for training | ||||||||
| SOC 2 and ISO 27001 | ||||||||
| HIPAA BAA | ||||||||
| FedRAMP | ||||||||
| Copyright protection | ||||||||
| Public code filter |
- Yes
- Partly
- No
- Not documented
- Words in cells are each vendor’s own name for the feature. Click a cell for its source.
Can it connect to tools through the Model Context Protocol (MCP)? Connects to MCP servers for tools, databases and APIs; admins can control which servers are allowed.
Claude Code can connect to hundreds of external tools and data sources through the Model Context Protocol (MCP) , an open source standard for AI-tool integrations.
Can it connect to tools through the Model Context Protocol (MCP)? MCP servers connect on web, desktop, CLI and IDE; admins control plugins and connectors.
Model Context Protocol (MCP) lets ChatGPT and Codex use tools and information from other systems.
Can it connect to tools through the Model Context Protocol (MCP)? Connects to local and remote MCP servers in the editor, CLI and cloud agents.
Model Context Protocol (MCP) enables Cursor to connect to external tools and data sources.
Can it connect to tools through the Model Context Protocol (MCP)? Connects to MCP servers over stdio, SSE or HTTP; admins can add them for the whole enterprise.
MCP is an open protocol that enables Devin to use hundreds of external tools and data sources.
Can it connect to tools through the Model Context Protocol (MCP)? MCP works in IDEs, the CLI, the app and cloud agent; on Business and Enterprise an admin policy must allow it.
MCP works across all major Copilot surfaces—whether you're working in an IDE, using GitHub Copilot CLI, working in the GitHub Copilot app, or delegating tasks to an agent on GitHub.com.
Can it connect to tools through the Model Context Protocol (MCP)? Agents can use MCP servers; admins can turn MCP on or off and set an allowlist of servers.
You can configure access to Model Context Protocol (MCP) servers, which lets AI agents securely access additional data sources and tools.
Can it connect to tools through the Model Context Protocol (MCP)? Connects to MCP servers; admins on IDE Services or JetBrains Central can preset and lock them.
AI Assistant can interact with external tools and data sources through the Model Context Protocol (MCP)
Can it connect to tools through the Model Context Protocol (MCP)? Connects to local and remote MCP servers; admins can govern which servers are allowed.
Model Context Protocol (MCP) extends Kiro's capabilities by connecting to specialized servers that provide additional tools and context.
Can people or builders choose models from more than one AI company? Claude models only, though they can be served through Amazon Bedrock, Google Cloud or Microsoft Foundry.
Anthropic doesn’t endorse, maintain, or audit third-party gateway products, and doesn’t support routing Claude Code to non-Claude models through any gateway.
Can people or builders choose models from more than one AI company? ChatGPT plans give OpenAI models only; the local CLI can run open models through Ollama or LM Studio.
Codex can run against a local “open source” provider such as Ollama or LM Studio when you pass --oss .
Can people or builders choose models from more than one AI company? Models from OpenAI, Anthropic, Google and SpaceXSI (Grok), plus Cursor's own; third-party models cost extra.
Cursor supports frontier models from OpenAI, Anthropic, Google, SpaceXSI, and more.
Can people or builders choose models from more than one AI company? Choose models from OpenAI, Anthropic, Google, xAI and Cognition's own SWE models; admins set an allowlist.
Tour Devin Desktop, the Agent Development Environment for local and cloud agents, with support for ACP, all frontier models, and leading open models.
Can people or builders choose models from more than one AI company? Models from OpenAI, Anthropic, Google, xAI and Microsoft; admins enable or disable each model.
We are bringing developer choice to GitHub Copilot with Anthropic's Claude 3.5 Sonnet, Google's Gemini 1.5 Pro, and OpenAI's o1-preview and o1-mini.
Can people or builders choose models from more than one AI company? Gemini models by default; admins can switch on Anthropic Claude models, billed by tokens.
Antigravity in Gemini Enterprise supports included first-party Gemini models and administrator-enabled third-party partner models
Can people or builders choose models from more than one AI company? JetBrains AI includes OpenAI, Google and Anthropic models; teams can also bring their own keys or local models.
providers of the AI models connected to JetBrains AI, such as OpenAI, Google, and Anthropic
Can people or builders choose models from more than one AI company? Pick OpenAI, Anthropic or open-weight models, or let Auto route each task.
Kiro gives you access to frontier and open weight AI models from OpenAI, Anthropic, and other providers.
Can it plan and make a change across many files from inside the editor? Reads the codebase, edits many files and runs commands, from the VS Code or JetBrains extension or the terminal.
It understands your entire codebase and can work across multiple files and tools to get things done.
Can it plan and make a change across many files from inside the editor? The IDE extension edits files in the open project and shows diffs; longer work can move to the cloud.
Bring open files and selections into the prompt, review edits in place, and hand off longer work without breaking your flow.
Can it plan and make a change across many files from inside the editor? Agent searches the codebase, edits files and runs commands inside the Cursor editor.
Agent is Cursor's assistant that can complete complex coding tasks independently, run terminal commands, and edit code.
Can it plan and make a change across many files from inside the editor? The Devin Local agent in Devin Desktop can plan first, then edit files across the workspace.
Plan mode helps you think through complex tasks by developing a detailed implementation plan before writing any code.
Can it plan and make a change across many files from inside the editor? Agent mode picks the files, edits them and runs commands in the editor; admins can turn it off.
In agent mode, Copilot takes a high-level task, decides which files to change, makes the edits, and runs commands as needed, iterating until the task is done.
Can it plan and make a change across many files from inside the editor? Agents in the app, CLI and IDE extensions plan and make multi-step changes such as features and refactors.
Developers can delegate multi-step software engineering tasks—such as implementing new features, refactoring legacy modules, migrating APIs, and generating comprehensive test suites—directly to autonomous AI agents.
Can it plan and make a change across many files from inside the editor? Junie plans and carries out multi-step changes across the project from inside the IDE.
Junie is an AI coding agent developed by JetBrains whose primary task is to autonomously plan and execute complex, multi-step actions based on your prompt.
Can it plan and make a change across many files from inside the editor? Turns a request into a written spec, then plans and makes the change across the codebase.
Instead of reacting to individual prompts, Kiro works from structured specifications to plan, implement, and verify changes across your entire codebase.
Can it take a task, work on it in the cloud on its own and open a pull request? Runs in an Anthropic-managed VM and pushes a branch; you open the PR from the session.
Connect a GitHub repository, submit a task, and review the PR without local setup.
Can it take a task, work on it in the cloud on its own and open a pull request? Tasks run in their own cloud workspace on a GitHub or GitLab repo; you open the pull request when ready.
Inspect changed files and check results, request follow-up changes, and commit or open a pull request when you're ready.
Can it take a task, work on it in the cloud on its own and open a pull request? Agents run in isolated cloud VMs, work on a branch and open pull requests; an admin connects source control first.
The agent can inspect the full workspace, make coordinated changes, and open pull requests in the repos it changes.
Can it take a task, work on it in the cloud on its own and open a pull request? Devin Cloud runs each task on its own VM and opens pull requests in your repositories.
Integrating Devin with your GitHub organization enables Devin to create pull requests, respond to PR comments, and collaborate directly within your repositories.
Can it take a task, work on it in the cloud on its own and open a pull request? Delegated tasks run in a GitHub cloud sandbox or Actions runner and come back as a pull request.
The agent works in a cloud environment and can research a repository, create a plan, make changes on a branch, and open a pull request for review.
Can it take a task, work on it in the cloud on its own and open a pull request? Remote Agents run in Google's cloud only for personal AI Pro and Ultra accounts, not business accounts.
It isn’t available on the free plan, on Google AI Plus, or for business accounts (Google Workspace or Google Cloud).
Can it take a task, work on it in the cloud on its own and open a pull request? Junie runs on your GitHub runners and opens PRs from issues; Air cloud agents need JetBrains Central.
Junie creates pull requests with code changes or fixes from issue descriptions, PR reviews, or comments.
Can it take a task, work on it in the cloud on its own and open a pull request? Runs tasks in cloud sandboxes and opens PRs on GitHub or GitLab; an admin must turn on Cloud Sessions.
Connect your GitHub or GitLab repositories or both, describe what you need, and Kiro writes code, coordinates changes across repos, and opens pull requests.
Can it review a pull request and leave comments on it? Team and Enterprise, GitHub only, billed on usage on top of seats; not available with ZDR or HIPAA setup.
Findings arrive as inline comments on the lines concerned, ranked by severity , with a summary and a check run on the pull request.
Can it review a pull request and leave comments on it? Reviews GitHub pull requests on request (@codex review) or automatically; GitLab support is in preview.
Codex posts a review on the pull request, just like a teammate would.
Can it review a pull request and leave comments on it? Bugbot reviews pull requests on GitHub, Bitbucket and Azure DevOps and comments with suggested fixes.
Bugbot analyzes PR diffs and leaves comments with explanations and fix suggestions.
Can it review a pull request and leave comments on it? Reviews GitHub and GitLab pull requests on request or automatically and can post findings as comments.
Devin Review groups related changes in a pull request, explains them alongside the diff, and points out potential bugs and security issues.
Can it review a pull request and leave comments on it? Copilot leaves review comments with suggested fixes; reviews can run automatically on every pull request.
Copilot's review comments work like comments from human reviewers.
Can it review a pull request and leave comments on it? Agents can review code changes and pull requests; no bot that posts PR comments is documented for Antigravity.
Autonomous agents analyze code changes and pull requests to detect bugs, security vulnerabilities, and style violations, offering actionable suggestions and automated fixes before code is merged.
Can it review a pull request and leave comments on it? A GitHub workflow reviews each opened or updated PR and posts a summary and inline comments.
When done, Junie adds a PR comment with a review summary and posts inline comments with suggestions directly to the PR.
Can it review a pull request and leave comments on it? Kiro CLI can run code reviews in a CI pipeline; no built-in bot that posts PR comments is documented.
Headless mode lets you run Kiro CLI as part of your CI/CD pipeline to automate code reviews, generate tests, or troubleshoot build failures
Is there a version that works from the command line? The terminal CLI is the main surface and the most complete one.
The full-featured CLI for working with Claude Code directly in your terminal.
Is there a version that works from the command line? Interactive terminal agent, plus codex exec for scripts and CI.
Inspect code, make changes, run commands, and automate repeatable work without leaving your terminal.
Is there a version that works from the command line? Interactive terminal agent plus a print mode for scripts and CI.
Cursor CLI lets you interact with AI agents directly from your terminal to write, review, and modify code.
Is there a version that works from the command line? Devin CLI is a local terminal agent that can also hand tasks to Devin Cloud.
Get up and running in 2 minutes with Devin CLI, a local command-line coding agent with deep Devin Cloud integration.
Is there a version that works from the command line? Copilot CLI is included in all plans; for Business and Enterprise an org policy must enable it.
The command-line interface (CLI) for GitHub Copilot allows you to use Copilot directly from your terminal.
Is there a version that works from the command line? The Antigravity CLI (agy) works with Gemini Enterprise licences, including headless runs.
Enterprise integration is supported for the Antigravity App, Antigravity CLI, and Antigravity IDE extensions.
Is there a version that works from the command line? Junie CLI runs the agent in the terminal on Linux, macOS and Windows.
Junie CLI is the agentic coding tool by JetBrains that provides an interactive terminal interface for developers to review, write, and modify code.
Is there a version that works from the command line? Kiro CLI runs the same agent in the terminal, including headless runs in CI.
Kiro CLI is a tool that allows you to interact with Kiro’s AI agents directly from your terminal.
Does it work inside both VS Code and JetBrains IDEs? VS Code extension with its own panel; the JetBrains plugin runs the CLI inside the IDE with IDE diffs.
Claude Code integrates with JetBrains IDEs through a dedicated plugin, providing features like interactive diff viewing, selection context sharing, and more.
Does it work inside both VS Code and JetBrains IDEs? Extension for VS Code and its forks; JetBrains IDEs and Xcode offer Codex through their own AI chat.
VS Code and compatible editors use the Codex extension; Xcode and JetBrains IDEs provide their own integrations.
Does it work inside both VS Code and JetBrains IDEs? Works in JetBrains IDEs through ACP, but not inside VS Code: Cursor is its own separate editor.
Use Cursor's AI agent in IntelliJ IDEA, PyCharm, WebStorm, and other JetBrains IDEs through the Agent Client Protocol (ACP).
Does it work inside both VS Code and JetBrains IDEs? Chat plugins for VS Code and JetBrains; the agent runs in JetBrains over ACP. The old JetBrains plugin is in maintenance.
Chat and its related features are only supported in: VS Code, JetBrains IDEs, Eclipse, Xcode, and Visual Studio.
Does it work inside both VS Code and JetBrains IDEs? Agent mode runs in VS Code, Visual Studio, JetBrains IDEs, Xcode and Eclipse.
Agent mode is available in Visual Studio Code, Visual Studio, JetBrains IDEs, Xcode, and Eclipse.
Does it work inside both VS Code and JetBrains IDEs? Extensions for VS Code, Visual Studio, JetBrains IDEs, Zed and Xcode.
Supported IDEs: Visual Studio Code, Visual Studio, JetBrains IDEs (such as IntelliJ IDEA, PyCharm, WebStorm, and GoLand), Zed, and Xcode.
Does it work inside both VS Code and JetBrains IDEs? Built into JetBrains IDEs; the 2025 VS Code preview is no longer in the current docs.
AI Assistant is a collection of AI-powered features and coding agents integrated into JetBrains IDEs.
Does it work inside both VS Code and JetBrains IDEs? Works in JetBrains through ACP; for VS Code users Kiro is its own VS Code-based IDE, not an extension.
This means you can use Kiro's agentic capabilities in JetBrains IDEs, Zed, and other ACP-compatible editors.
Does it support single sign-on (SAML/OIDC) and automatic user provisioning (SCIM)? SSO on Team and Enterprise; SCIM provisioning on Enterprise only.
Claude Enterprise includes SSO, domain capture, SCIM provisioning, role-based permissions, audit logs, custom data retention, the Compliance API, and an option for US-only inference.
Does it support single sign-on (SAML/OIDC) and automatic user provisioning (SCIM)? SAML SSO is in Business; SCIM provisioning is Enterprise only.
Enterprise-level security and controls, including SCIM, EKM, user analytics, domain verification, and role-based access control
Does it support single sign-on (SAML/OIDC) and automatic user provisioning (SCIM)? SAML SSO is included on Teams and Enterprise; SCIM provisioning is Enterprise only.
SCIM 2.0 provisioning automatically manages your team members and directory groups through your identity provider.
Does it support single sign-on (SAML/OIDC) and automatic user provisioning (SCIM)? SAML or OIDC SSO on Teams and Enterprise; SCIM provisioning and IdP groups are Enterprise only.
Auto-membership, IdP groups, role assignment, and SCIM are Enterprise only.
Does it support single sign-on (SAML/OIDC) and automatic user provisioning (SCIM)? SAML SSO, or managed user accounts provisioned from your identity provider with SCIM.
This gives you access to enterprise-grade integrations with identity providers for authentication and provisioning, without needing to pay for GitHub Enterprise licenses.
Does it support single sign-on (SAML/OIDC) and automatic user provisioning (SCIM)? Sign-in through Google Cloud or an external IdP such as Okta; no SCIM statement for Antigravity.
Bring Your Own Identity (BYOID) uses Workforce Identity Federation (WIF) to let your organization authenticate through an external identity provider, such as Okta, instead of a standard Google Account
Does it support single sign-on (SAML/OIDC) and automatic user provisioning (SCIM)? SAML or OAuth SSO and Okta/Entra user sync come with IDE Services (AI Enterprise); SCIM is not named.
IDE Services provides smooth integration with Okta, Microsoft Entra ID, Google, and other third-party authentication providers.
Does it support single sign-on (SAML/OIDC) and automatic user provisioning (SCIM)? SSO through IAM Identity Center, Okta or Entra ID, with SCIM to sync users and groups.
When using Okta, you need to create two applications: one OIDC application for user login, and one SCIM provisioning application
Can admins export or query a log of who did what? Claude Code activity is exported as OpenTelemetry events to your own SIEM; admins set up the export.
Every event carries identity attributes that tie tool calls, MCP activity, and permission decisions back to the user who triggered them.
Can admins export or query a log of who did what? Enterprise only; exports Codex activity records to SIEM or investigation tools.
Audit logs and usage monitoring via the Compliance API
Can admins export or query a log of who did what? Enterprise only; logs admin and account changes with actor and IP, but never prompts or generated code.
Audit logs record who changed what in your team and organization: sign-ins, membership, roles, API keys, settings, integrations, Cloud Agent environments, Grok Bot configuration, and Origin repositories.
Can admins export or query a log of who did what? Enterprise admins can pull enterprise and organization audit logs through the Devin API.
For cross-organization management — analytics, audit logs, user management, billing, and infrastructure.
Can admins export or query a log of who did what? Logs Copilot settings, licence and agent events for 180 days and can stream to a SIEM; prompts are not logged.
You can use the audit log to review actions taken in your enterprise.
Can admins export or query a log of who did what? Admins can log prompts and responses to their own Google Cloud project as an audit trail; off by default.
Enable this setting to record interactions between developers and the AI to compute usage metrics and maintain an audit trail for security reviews.
Can admins export or query a log of who did what? Air Governance logs admin changes with who and when, and admins can export it; AI sessions are in Session explorer.
You can review changes and see who made them, when they happened, and whether they succeeded.
Can admins export or query a log of who did what? Admins can log users' prompts and responses to their own S3 bucket; CloudTrail records Kiro tool calls.
Prompt logs provide you with a record of all the prompts that users enter into Kiro chat in their IDE or the CLI.
Can admins see who is using it and how much? Admins see active users, sessions, accepted lines and a leaderboard; per-user spend via export.
Claude Code provides analytics dashboards to help organizations understand developer usage patterns, track contribution metrics, and measure how Claude Code impacts engineering velocity.
Can admins see who is using it and how much? Admin Console shows Codex usage, credits and active users; an Analytics API gives aggregated reports.
Insights in the Admin Console helps you understand what teams use ChatGPT Work and Codex for and where credits are spent.
Can admins see who is using it and how much? Dashboard shows usage per member with CSV export; the analytics API is Enterprise only.
Team admins have access to data for themselves and all other users in the team.
Can admins see who is using it and how much? Admins see consumption by organization and by user; Desktop also has per-group analytics.
Org admins can break it down by user from Consumption Analytics inside Organization Settings.
Can admins see who is using it and how much? Dashboards plus APIs and exports with per-user and per-repository usage.
The APIs provide enterprise- and organization-scoped reports with repository- and user-level detail.
Can admins see who is using it and how much? Metadata logging tracks usage across the organization, including third-party model tokens.
Enable this setting to record product metadata to help you track usage and analytics across your organization, including third-party model token consumption
Can admins see who is using it and how much? Admins see active AI users, credit use and who hit their limits, filtered by tool.
The AI adoption and usage segment lets you track AI adoption and engagement in your organization over a period of time by providing insight into the active use of AI in the organization.
Can admins see who is using it and how much? An admin dashboard plus daily per-user activity reports to S3 or OpenTelemetry.
User activity reports show you what individual users are up to in Kiro.
Can the company choose the region where its data is stored? Usage-based Enterprise only, US only, and it covers inference, not where data is stored; usage costs 1.1x.
When the setting is on, all inference for your organization runs on servers located in the United States.
Can the company choose the region where its data is stored? New Enterprise customers only; covers the Codex app and CLI but not Codex Web cloud tasks.
Eligible API customers and new ChatGPT Enterprise/Edu customers can choose to have in-scope customer content stored at rest in supported countries.
Can the company choose the region where its data is stored? US-only residency for enrolled customers; EU and other regions are still in development.
Today, customers can enroll in US-only data residency (inference, processing, and storage in the US).
Can the company choose the region where its data is stored? A Frankfurt cluster keeps Desktop data in the EU for enterprises; no region choice is documented for Devin Cloud.
Europe cluster: A datacenter in Frankfurt for our European enterprise customers that guarantees that all data processing and data retention stays within EU borders.
Can the company choose the region where its data is stored? Only on GitHub Enterprise Cloud with data residency (GHE.com), US or EU only, and it costs 10% more AI credits.
If your enterprise uses GitHub Enterprise Cloud with data residency, you can enable a policy to ensure that all inference processing and associated data for GitHub Copilot remain within your designated geographic region .
Can the company choose the region where its data is stored? Choose US or EU multi-region endpoints; image generation runs only on global endpoints.
Antigravity CLI, the Antigravity App, and IDE extensions support multi-region deployment endpoints to satisfy regional data residency requirements
Can the company choose the region where its data is stored? Only by self-hosting AI Enterprise; no region choice is documented for the JetBrains AI cloud service.
AI Enterprise runs on-premises as part of JetBrains IDE Services, ensuring complete control over data and AI operations within your organization’s infrastructure.
Can the company choose the region where its data is stored? Data stays in the profile's region: US East, Frankfurt or GovCloud. Some models may run in other regions.
your content, such as prompts and responses, may be stored in the region where your profile is configured in order to provide and maintain the service
Is business data excluded from model training by default? Team and Enterprise code and prompts are not used for training unless the company opts in.
Anthropic does not train generative models using code or prompts sent to Claude Code under commercial terms
Is business data excluded from model training by default? Business and Enterprise data is not used for training by default.
No training on your business data by default.
Is business data excluded from model training by default? Privacy Mode is on by default for team members and admins can lock it on.
For teams, Privacy Mode is enabled by default for all team members.
Is business data excluded from model training by default? Enterprise data is never used without written consent; on Teams, training is on until an admin turns it off.
If you are an Enterprise customer, we will never train on your data without your express prior written consent.
Is business data excluded from model training by default? Business and Enterprise data is not used for training; individual plans may be, unless the user opts out.
GitHub does not use Copilot Business or Copilot Enterprise customer data to train AI models.
Is business data excluded from model training by default? Under Google Cloud terms, code and prompts are not used to train Google's base models.
All agent activity remains strictly inside your enterprise boundary under Google Cloud Terms of Service—your code is never used to train base models.
Is business data excluded from model training by default? Off by default for organizations; an admin must allow it before users can share code data.
By default, AI tools and features will not share any code-related data, and your users cannot opt in.
Is business data excluded from model training by default? Enterprise content is not used; free and individual plans may be used unless they opt out.
We do not use content from Kiro enterprise users for service improvement.
Does the vendor state SOC 2 Type II and ISO 27001 for this product? SOC 2 Type 2 report and ISO 27001 certificate are available from Anthropic's Trust Center.
Learn more and access resources (SOC 2 Type 2 report, ISO 27001 certificate, etc.) at Anthropic Trust Center
Does the vendor state SOC 2 Type II and ISO 27001 for this product? SOC 2 Type 2 covers ChatGPT business products and ISO 27001 covers ChatGPT Enterprise; Codex is not named.
OpenAI has undergone an independent SOC 2 Type 2 examination of controls relevant to Security, Availability, Confidentiality, and Privacy for its API and ChatGPT business product services.
Does the vendor state SOC 2 Type II and ISO 27001 for this product? SOC 2 Type II, ISO 27001 and ISO 42001; reports on request from the trust center.
Cursor holds AIUC-1, ISO/IEC 27001:2022, and ISO/IEC 42001:2023 certifications, along with a SOC 2 Type II attestation.
Does the vendor state SOC 2 Type II and ISO 27001 for this product? SOC 2 Type II is stated; no ISO 27001 statement found in Cognition's docs.
Cognition has been SOC 2 Type II certified since September 2024.
Does the vendor state SOC 2 Type II and ISO 27001 for this product? In GitHub ISO 27001 scope since 2024; a Dec 2024 changelog adds them to SOC 2 Type II reports.
Copilot Business and Copilot Enterprise are now included in the scope of GitHub's Information Security Management System, as reflected in our ISO 27001 certificate updated on May 9, 2024.
Does the vendor state SOC 2 Type II and ISO 27001 for this product? Google lists ISO 27001 and SOC 2 among the certifications Antigravity in Gemini Enterprise does not have.
Antigravity in Gemini Enterprise doesn't support the following compliance certifications and security controls: Access Transparency (AXT) FedRAMP (Moderate and High) Impact Level 4 (IL4) and Impact Level 5 (IL5) International Organization for Standardization (ISO) certifications, including ISO 27001 and ISO 42001
Does the vendor state SOC 2 Type II and ISO 27001 for this product? SOC 2 Type II covers Air Governance; no ISO 27001 statement found for JetBrains AI.
SOC 2 Type II certification covers Air Governance and the underlying governance layer.
Does the vendor state SOC 2 Type II and ISO 27001 for this product? Kiro is in AWS's ISO 27001 scope; AWS also lists Kiro in its SOC report scope.
Kiro is included in the defined scope of AWS's ISO/IEC 27001:2022 certification.
Will the vendor sign a HIPAA business associate agreement that covers this product? Enterprise only; the BAA covers the CLI and Desktop Code tab, not cloud sessions, IDE extensions or Code Review.
The HIPAA configuration is an organization setting on Claude Enterprise plans, for organizations that handle protected health information (PHI) and have a Business Associate Agreement (BAA) with Anthropic.
Will the vendor sign a HIPAA business associate agreement that covers this product? Only for ChatGPT for Healthcare, Clinicians or Regulated workspaces with a BAA, not standard Business.
If you use ChatGPT for Healthcare, ChatGPT for Clinicians, or a Regulated workspace, have an applicable OpenAI Business Associate Agreement (BAA), and have the required Codex access, OpenAI handles PHI it receives from Codex consistent with the BAA.
Will the vendor sign a HIPAA business associate agreement that covers this product? Enterprise only, and the BAA covers only listed Eligible Services with Privacy Mode locked on.
Cursor supports HIPAA Business Associate Agreements (BAAs) for Enterprise customers.
Will the vendor sign a HIPAA business associate agreement that covers this product? Searched Devin docs, security pages and platform terms; no HIPAA or BAA statement.
No statement from Cognition on its own pages. Searched Devin docs, security pages and platform terms; no HIPAA or BAA statement.
Will the vendor sign a HIPAA business associate agreement that covers this product? Searched GitHub Docs, customer terms, plans and trust pages; no statement on a HIPAA BAA for Copilot.
No statement from GitHub on its own pages. Searched GitHub Docs, customer terms, plans and trust pages; no statement on a HIPAA BAA for Copilot.
Will the vendor sign a HIPAA business associate agreement that covers this product? Antigravity in Gemini Enterprise is on the list of products covered by the Google Cloud BAA.
Antigravity in Gemini Enterprise
Will the vendor sign a HIPAA business associate agreement that covers this product? Searched JetBrains AI, Air and IDE Services docs and the JetBrains AI terms; no HIPAA or BAA statement.
No statement from JetBrains on its own pages. Searched JetBrains AI, Air and IDE Services docs and the JetBrains AI terms; no HIPAA or BAA statement.
Will the vendor sign a HIPAA business associate agreement that covers this product? IDE and CLI are HIPAA eligible, so covered by the AWS BAA; Kiro Web is not listed.
Kiro IDE and CLI are HIPAA eligible.
Is this product FedRAMP authorized for US government use? FedRAMP High only via Claude for Government, a separate agency offering, not commercial plans.
Claude and Claude Code are live, built on the same application our commercial customers use and delivered through a FedRAMP High authorized environment.
Is this product FedRAMP authorized for US government use? ChatGPT Enterprise is FedRAMP Moderate; there Codex runs only through the desktop app and CLI with a FedRAMP setup.
OpenAI offers specific FedRAMP configurations version of ChatGPT Desktop and Codex via the CLI.
Is this product FedRAMP authorized for US government use? Searched Cursor docs, security page and enterprise compliance docs; no FedRAMP authorization stated.
No statement from Anysphere on its own pages. Searched Cursor docs, security page and enterprise compliance docs; no FedRAMP authorization stated.
Is this product FedRAMP authorized for US government use? Devin Desktop and CLI are FedRAMP High authorized; Devin Cloud and Devin Review are still in process.
Devin Desktop and CLI are already FedRAMP High Authorized and compliant for workloads with ITAR and DoW IL4, IL5, IL6 requirements.
Is this product FedRAMP authorized for US government use? Admins can limit Copilot to FedRAMP Moderate model hosts on US GHE.com; Copilot itself is not yet.
Copilot itself will become FedRAMP authorized as part of GHEC-DR's authorization path .
Is this product FedRAMP authorized for US government use? Not with Gemini Enterprise; the app and CLI can run in a FedRAMP High boundary only via Assured Workloads and the API.
Antigravity in Gemini Enterprise doesn't support the following compliance certifications and security controls: Access Transparency (AXT) FedRAMP (Moderate and High)
Is this product FedRAMP authorized for US government use? Searched JetBrains AI, Air and IDE Services docs; no FedRAMP statement.
No statement from JetBrains on its own pages. Searched JetBrains AI, Air and IDE Services docs; no FedRAMP statement.
Is this product FedRAMP authorized for US government use? IDE and CLI are FedRAMP High in AWS GovCloud (US); each model needs its own authorization.
Kiro is now authorized for FedRAMP High and Department of Defense Cloud Computing Security Requirements Guide Impact Levels 4 and 5 in AWS GovCloud (US).
Does the vendor defend customers if AI output is claimed to infringe copyright? Commercial Terms (Team, Enterprise): Anthropic defends IP claims over paid use and outputs.
means a third-party claim, suit, or proceeding alleging that Customer’s paid use of the Services (which includes data Anthropic has used to train a model that is part of the Services) in accordance with these Terms or Outputs generated through such authorized use violates any third-party intellectual property right.
Does the vendor defend customers if AI output is claimed to infringe copyright? Covers output for Enterprise, Edu and Healthcare, not Business; lapses if filters are disabled.
OpenAI’s indemnification obligations to Enterprise customers under the Agreement include claims that Customer’s use or distribution of Output infringes a third party’s intellectual property right.
Does the vendor defend customers if AI output is claimed to infringe copyright? In the MSA used for Enterprise; self-serve terms have no such defense, and it lapses if filters are disabled.
Anysphere shall indemnify, defend, and hold harmless Customer, its Affiliates, and each of their officers, directors, employees, and agents from and against any and all claims, lawsuits, losses, damages, liabilities, and costs (including reasonable attorneys' fees) (" Losses ") arising out of any third-party claim, suit, action, or proceeding (" Third-Party Claim ") that the Service or any Suggestion infringes or misappropriates such third party's intellectual property rights.
Does the vendor defend customers if AI output is claimed to infringe copyright? Paid plans: Cognition defends claims that the service infringes copyright; the terms do not mention AI output.
Cognition shall defend Customer against any third-party claim, suit, action, or proceeding (“Third-Party Claim”) that the Services, when used in accordance with the terms of this Agreement, infringes or misappropriates such third party's patents, copyrights, or trade secrets
Does the vendor defend customers if AI output is claimed to infringe copyright? The IP defense in GitHub's customer agreement extends to AI outputs; Microsoft's required mitigations apply.
If your Agreement provides for the defense of third party claims, that provision will apply to your use of Generative AI Services, including to Outputs.
Does the vendor defend customers if AI output is claimed to infringe copyright? Gemini Enterprise is on Google's indemnified list; Antigravity and Claude models are not named separately.
The following Services are covered by the Additional Google Indemnification Obligations in the Service Specific Terms.
Does the vendor defend customers if AI output is claimed to infringe copyright? The JetBrains AI terms make the customer indemnify JetBrains for claims about outputs, not the reverse.
This includes any allegation that Your Inputs and Outputs, or their use, development, design, production, advertising, or marketing, infringe upon the rights of someone else
Does the vendor defend customers if AI output is claimed to infringe copyright? Paid subscribers are covered under AWS Service Terms section 50.10; some models are excluded.
Yes, we provide an output indemnity for paid Kiro subscribers.
Can admins block or flag suggestions that match public code? Searched Claude Code docs (security, admin, legal) and Anthropic terms; no public-code filter.
No statement from Anthropic on its own pages. Searched Claude Code docs (security, admin, legal) and Anthropic terms; no public-code filter.
Can admins block or flag suggestions that match public code? Searched Codex docs, admin and security guides, and OpenAI service terms; no public-code matching filter.
No statement from OpenAI on its own pages. Searched Codex docs, admin and security guides, and OpenAI service terms; no public-code matching filter.
Can admins block or flag suggestions that match public code? Searched Cursor docs, enterprise safety docs and terms; no built-in public-code filter. Docs suggest custom hooks.
No statement from Anysphere on its own pages. Searched Cursor docs, enterprise safety docs and terms; no built-in public-code filter. Docs suggest custom hooks.
Can admins block or flag suggestions that match public code? Enterprise only and off by default: blocks code that matches public code and makes Devin rewrite it.
When a match against public code is detected, the write is blocked and Devin rewrites the implementation instead.
Can admins block or flag suggestions that match public code? Policy blocks suggestions that match public GitHub code, or shows them with the source and licence.
Depending on the policy that applies to your account or organization, matching suggestions can be blocked or shown with information about the matching code.
Can admins block or flag suggestions that match public code? Searched Antigravity docs and Gemini Enterprise developer tool settings; no filter for code that matches public code.
No statement from Google on its own pages. Searched Antigravity docs and Gemini Enterprise developer tool settings; no filter for code that matches public code.
Can admins block or flag suggestions that match public code? Searched AI Assistant, Junie and Air docs; no filter for suggestions that match public code.
No statement from JetBrains on its own pages. Searched AI Assistant, Junie and Air docs; no filter for suggestions that match public code.
Can admins block or flag suggestions that match public code? Code matching public code is off by default; admins can keep it off and users see a reference log.
Administrators can opt out of code references in the Kiro console.
Latest change, 9 Oct 2026: Google Antigravity. Google stops selling new Gemini Code Assist Standard and Enterprise subscriptions. All changes
Who sells what
A few companies sell in most of these categories, so start from the row you already pay for. Everyone else sells one kind.